Smart Load Balancing in Cloud Environments Using Artificial Intelligence

Abstract

This study evaluates the performance of AI-powered load balancing in cloud environments, comparing it with traditional load balancing techniques such as Round Robin, Least Connection, and Static Threshold. The AI-based approach demonstrated superior performance, achieving the lowest response time (162 ms) and highest throughput (1045 requests/sec). Scalability tests showed that the AI model maintained stable performance even as workload increased, in contrast to traditional methods, which exhibited significant performance degradation under high traffic. Additionally, AI-based load balancing achieved better resource utilization, reduced energy consumption, and improved fault detection accuracy (95%) with faster recovery times. These results suggest that AI-driven techniques can effectively enhance cloud system performance by dynamically adapting to changing workloads, ensuring higher efficiency, scalability, and reliability. This study highlights the potential of AI in optimizing cloud resource management and mitigating the limitations of conventional load balancing methods.